Get sort query parameter to work again

[evz]

As mentioned in #51, sorting the API responses doesn't work. When one digs into why it was failing, it had to do with the way that Django was attempting to construct the DISTINCT ON clause for PostgreSQL. Since PostgreSQL requires that the ORDER BY portion of the query have fields in the same order as the fields in the DISTINCT ON portion of the query, and Imago just appeared to be constructing the query one step at a time, Django didn't really know what to do in that situation.

This PR just removes the DISTINCT ON thing since there really shouldn't be a situation where there is more than one object with the same ID. Does this seem sane?

[jpmckinney]

Can you split out the following into separate PRs:

• auth helper fix
• schema fix
• exposing membership contact details

[jpmckinney]

Hmm, nevermind - (some of) those commits already exist in master, but for some reason show up in the PR.

[jpmckinney]

@paultag added that line in https://github.com/opencivicdata/imago/commit/d977bba960c9596da2dafecf596f6d3338b08d92 so I assume there's a reason for it.

[evz]

@jpmckinney Yeah, I was kinda scratching my head about that, too. As far as I could tell, I was keeping my fork synced with opencivicdata.

[evz]

@jpmckinney I figure it's there for a reason. It's just causing Django to not know what to do when you're wanting to sort. I was hoping we could get to the bottom of it here.

[jpmckinney]

I mentioned paultag, since he would know what's going on - but I figure the preceding lines must sometimes return duplicates.

[fgregg]

ping @paultag

[fgregg]

@jpmckinney, here's the word from paultag

11:45 <+paultag> I'm afraid I can't gain context at the moment, but it's been 
                 on my queue
11:45 <+paultag> if my actions looked irrational, ignore them and revert

[jpmckinney]

Actually, can you rebase so that all the old commits don't show up here? I don't know what I'm commenting on.

[fgregg]

Replaced by #77